Business Modelling (8): Policy, Decisions and Control

How current performance is used to make strong, balanced decisions

An Extension to the Core principles and business models.

This short course is just one of several Extensions you can use on its own or add to other Dynamic Business models, and remember you can also buy the complete set here for a lower total price (opens in a new tab).

What will I learn?

This short course shows how to model "policies" that drive Decisions and control the system. (See the curriculum near the end of this page.) Most decisions we take have been made before - how many people to hire this month, how much to spend on marketing, what price to set, and so on. So we rarely have to make such decisions from first principles. Instead, we follow policies - or decision-making rules - that we think should work well. Many of these are never written down - they are just implicit - but they are there all the same. 

Unfortunately, easy and obvious policies can actually harm our business - setting spending rates as a % of revenue, or spending "last year + X%" for example. So we need to design better policies and test them before we use them for real. 

How does it work ?  Any policy - whether written down or not - consists of some common elements:

... a target for some indicator, often a performance result

... the current value for that same indicator

... a rule that works out what value to set for a decision, depending on the gap between the current value and the target

This class shows how to model decision-making policies as part of other models. Calculate the difference between a target and current value of some indicator, then set up a formula to calculate the next period's decision value.

The "RoI" of this course.

Well-designed policies ensure sustained strong performance - so careful design and testing is hugely valuable. Bad policies can be catastrophic, even destroying a business.  

Think ahead - and balance your objectives! The small IT support business shown here did well winning new clients and growing its profits ... but it had not pre-built the capacity of its staff to handle the extra business. By month 12, service quality was awful, existing clients were leaving and profits fell.  

Surely, management knew they would need more staff? Well, yes they did and their hiring policy would have told them to hire in advance, but another policy said to keep staff cost below X% of revenue - and that would only allow them to hire after the new clients had been won and started demanding service. 

A really BIG failed policy. On a totally different scale, a large factor in the 2008/9 financial crisis was the pursuit of "sub-prime mortgage" business by US and other countries' banks. The "policy" was to lend money to people with a high risk of being unable to repay, on the basis that the high interest charged would cover the losses to borrowers who defaulted. Unfortunately, while the principle could be OK - if questionably ethical - the banks failed to test the lending-rule policy adequately. The Canadian banks, meanwhile, mostly did not follow this flock of sheep over the cliff! The movie documentary "The Big Short" is not only great to watch, but actually portrays the dysfunctional policies that piled up on each other to trash the US economy!

Even for the small IT-support business, getting a very basic policy right turned out be worth tens of thousands of pounds each month (the excess profit during  ... months 9-15 was never sustainable)  ... all for the investment of a few hours in learning and modelling their policies.

Why are policies so hard to get right?

Delays occur, both in getting data about the current situation (it's not 'current' any more by the time we get it!), and in seeing the impact of any decision we take. So the world has already moved on and we are shooting at a moving target.  

Multiple decisions affect the same thing. The price we set and the promotions we do both affect our sales rate.  

A single decision affects more than one factor. That last price rise slowed our customer win-rate and lost us some customers and cut each customer's purchase rate.

Objectives conflict with each other. The price rise may have done those bad things to growth of customers and sales, but it sure increased our profits - but will that benefit last?!

Like all our Extensions models, you can get value from using this one in two ways. Test every set of related policies. Whether or not you model the wider business, it is useful to simulate any significant policy - or set of related policies - to see how they work under a range of circumstances. (Some policies are so well understood that we no longer leave them to humans alone - like the yield-management systems that set prices for airlines and theatres). But also include policies in overall planning models. As the simple case above shows, the knock-on effects of any decision can ripple through the whole enterprise, and as you can see from month-18 onwards, embedding a better policy can protect the business from trouble.   

Who is this course for?

Business leaders  Some heads of marketing, R&D, HR - even some CEOs, feel it is so important to get a grip on the quality of their business that they invest effort in modelling it themselves - even a top level appreciation of the model will aid your thinking.

Analysts and accountants. Every analyst supporting a business or functional leader should know how to model the quality of their organisation's resources,  and understand how they are changing through time so they can advise leaders on what needs to be done.

Consultants. These models are so powerful that clients will be blown away by the value you bring to their business. It's faster and easier than using spreadsheets and less prone to error. The  approach used provides a valuable addition to a consultants toolbox.

Business students (and teachers*) The mechanisms in this course are a critical element of how business works - so understanding them will give you an important additional method for approaching strategic questions - and employers will really value this skill.

* Teachers - we can provide materials from this course for you to use in your classes - contact us for information. Please do let us know where you teach and approximate student numbers in addition to any  questions or comments you have.

How long does it take?

Like all of our courses, you do not have to start at the beginning and slave through every lesson to the end - we suggest key items to focus on first. Working carefully through just those key lessons may take about a half-day.

You can then come back to fill in important extra principles and skills. That could be another ½ to 1 day, depending on how deeply you want to get into each lesson and practice it on your own cases – all time that you can claim as self-directed CPD (continuous professional development).

Then, of course, you can spend any time you want on consolidating your skills and following our detailed guidance on how to apply the modelling to your own case and issues. 

Why use these models - and when? 

Whether launching a new business, tackling a problem or opportunity, or planning for the longer term, surely you would simulate what could happen, just like with other complicated things we do. This lets us: 

Design what we might do before we risk serious cash and effort

Test those plans against all kinds of uncertainties (Would you rather experiment with the real world, using real money and people!?)

Manage the business or issue continually against a living model of what will likely happen next.
... and if you want to be really bold and "re-invent your business model" or undergo "digital transformation" it might be best to explore where you are going before you set out!

What is the alternative? 

Spreadsheets could theoretically calculate-out everything in a business, but no human could reliably capture a living business system with only rows and columns of numbers to work with.

Balanced scorecards transformed business control by looking beyond the financials! But they do not calculate-out what causes what across the business "machine", and they leave out key issues - like competition! There is no better BSC than a good DBM!

KPI systems often try to be more comprehensive than BSCs, but they often go too far, including every indicator and ratio anyone thinks important. There is no better KPI system than a good DBM!

Business-model canvas and tools like Strategyzer or Leanstack share the same limitations as balanced scorecards and spreadsheets.

Course Pricing

Note: If you purchase a single extensions class and then decide to upgrade to the full extensions course we provide a coupon at the end of of the individual classes which can be used to recover the cost of your single class.

  • Instant Access

    £75

    One off payment, no expiry to access

  • Enrol Today

Course Curriculum

Kim Warren

Kim is an experienced strategy professional, teacher and publisher of online courses and teaching resources on business modeling – fast becoming a main-stream capability for executives, consultants and business students. He also offers resources to help model non-business challenges, notably in health-care and international aid.

After senior corporate strategy roles, Kim joined London Business School, to teach on MBA and Executive programs. To overcome serious limitations with standard strategy methods, he developed the powerful strategy dynamics modelling method for designing and managing strategy for any organisation or challenge. Once a specialist skill, building these simulations is now easier, faster and more reliable than spreadsheet modeling. Such models mimic real-world behaviour and performance of businesses and other organisations with uncanny realism.

Kim is author of the prize-winning Competitive Strategy Dynamics (Wiley, 2002), a major strategy textbook Strategic Management Dynamics (Wiley, 2008), and summary e-book now widely used in MBA and executive teaching – Strategy Dynamics Essentials (Kindle, 2011). He is also co-founder of Strategy Dynamics Ltd, which publishes "serious games" and online courses exploiting the user-friendly modelling application, Silico.